Biblio

With the development of the Internet, the captcha technology has also been widely used. Captcha technology is used to distinguish between humans and machines, namely Completely Automated Public Turing test to tell Computers and Humans Apart. In this paper, an end-to-end deep CNN-RNN network model is constructed by studying the captcha recognition technology, which realizes the recognition of 4-character text captcha. The CNN-RNN model first constructs a deep residual convolutional neural network based on the residual network structure to accurately extract the input captcha picture features. Then, through the constructed variant RNN network, that is, the two-layer GRU network, the deep internal features of the captcha are extracted, and finally, the output sequence is the 4-character captcha. The experiments results show that the end-to-end deep CNN-RNN network model has a good performance on different captcha datasets, achieving 99% accuracy. And experiment on the few samples dataset which only has 4000 training samples also shows an accuracy of 72.9 % and a certain generalization ability.

With the advent of fifth generation (5G) network and increasingly powerful mobile devices, people can conveniently obtain network resources wherever they are and whenever they want. However, the problem of mobility support in current network has not been adequately solved yet, especially in inter-domain mobile scenario, which leads to poor experience for mobile consumers. MobilityFirst is a clean slate future Internet architecture which adopts a clean separation between identity and network location. It provides new mechanisms to address the challenge of wireless access and mobility at scale. However, MobilityFirst lacks effective ways to deal with multicast service over mobile networks. In this paper, we design an efficient multicast mechanism based on MobilityFirst architecture and present the deployment in current network at scale. Furthermore, we propose a hierarchical multicast packet header with additional destinations to achieve low-cost dynamic multicast routing and provide solutions for both the multicast source and the multicast group members moving in intra- or inter-domain. Finally, we deploy a multicast prototype system to evaluate the performance of the proposed multicast mechanism.

Based on the Lyapunov stability theory, a novel adaptive synchronization method is proposed for three chaotic systems with different orders. Then the proposed method is applied to secure communication. This paper designs a novel multistage chaotic synchronized secure communication system in which the encrypted information signal is transmitted to the receiver after two chaotic masking, and then recovered at the synchronized receiver. Numerical results show the success in transmitting a continuous signal and a discrete signal through three synchronized systems.

The purpose of this paper is to improve the safety of chaotic image encryption algorithm. Firstly, to achieve this goal, it put forward two improved chaotic system logistic and henon, which covered an promoted henon chaotic system with better probability density, and an 2-dimension logistic chaotic system with high Lyapunov exponents. Secondly, the chaotic key stream was generated by the new 2D logistic chaotic system and optimized henon mapping, which mixed in dynamic proportions. The conducted sequence has better randomness and higher safety for image cryptosystem. Thirdly, we proposed algorithm takes advantage of the compounded chaotic system Simulation experiment results and security analysis showed that the proposed scheme was more effective and secure. It can resist various typical attacks, has high security, satisfies the requirements of image encryption theoretical.

Generating adversarial samples is gathering much attention as an intuitive approach to evaluate the robustness of learning models. Extensive recent works have demonstrated that numerous advanced image classifiers are defenseless to adversarial perturbations in the white-box setting. However, the white-box setting assumes attackers to have prior knowledge of model parameters, which are generally inaccessible in real world cases. In this paper, we concentrate on the hard-label black-box setting where attackers can only pose queries to probe the model parameters responsible for classifying different images. Therefore, the issue is converted into minimizing non-continuous function. A black-box approach is proposed to address both massive queries and the non-continuous step function problem by applying a combination of a linear fine-grained search, Fibonacci search, and a zeroth order optimization algorithm. However, the input dimension of a image is so high that the estimation of gradient is noisy. Hence, we adopt a zeroth-order optimization method in high dimensions. The approach converts calculation of gradient into a linear regression model and extracts dimensions that are more significant. Experimental results illustrate that our approach can relatively reduce the amount of queries and effectively accelerate convergence of the optimization method.

Robust machine learning is currently one of the most prominent topics which could potentially help shaping a future of advanced AI platforms that not only perform well in average cases but also in worst cases or adverse situations. Despite the long-term vision, however, existing studies on black-box adversarial attacks are still restricted to very specific settings of threat models (e.g., single distortion metric and restrictive assumption on target model's feedback to queries) and/or suffer from prohibitively high query complexity. To push for further advances in this field, we introduce a general framework based on an operator splitting method, the alternating direction method of multipliers (ADMM) to devise efficient, robust black-box attacks that work with various distortion metrics and feedback settings without incurring high query complexity. Due to the black-box nature of the threat model, the proposed ADMM solution framework is integrated with zeroth-order (ZO) optimization and Bayesian optimization (BO), and thus is applicable to the gradient-free regime. This results in two new black-box adversarial attack generation methods, ZO-ADMM and BO-ADMM. Our empirical evaluations on image classification datasets show that our proposed approaches have much lower function query complexities compared to state-of-the-art attack methods, but achieve very competitive attack success rates.

Bitcoin [1] as a popular digital currency has been a target of theft and other illegal activities. Key to the forensic investigation is to identify bitcoin addresses involved in bitcoin transfers. This paper presents a framework, FABT, for forensic analysis of bitcoin transactions by identifying suspicious bitcoin addresses. It formalizes the clues of a given case as transaction patterns defined over a comprehensive set of features. FABT converts the bitcoin transaction data into a formal model, called Bitcoin Transaction Net (BTN). The traverse of all bitcoin transactions in the order of their occurrences is captured by the firing sequence of all transitions in the BTN. We have applied FABT to identify suspicious addresses in the Mt.Gox case. A subgroup of the suspicious addresses has been found to share many characteristics about the received/transferred amount, number of transactions, and time intervals.

Big data of mobile network privacy is vulnerable to clear text attack in the process of storage and mixed network information sharing, which leads to information leakage. Through the mixed encryption of data of mobile network privacy big data to improve the confidentiality and security of mobile network privacy big data, a mobile network privacy big data hybrid encryption algorithm based on hyperchaos theory is proposed. The hybrid encryption key of mobile network privacy big data is constructed by using hyperchaotic nonlinear mapping hybrid encryption technology. Combined with the feature distribution of mobile network privacy big data, the mixed encrypted public key is designed by using Logistic hyperchaotic arrangement method, and a hyperchaotic analytic cipher and block cipher are constructed by using Rossle chaotic mapping. The random piecewise linear combination method is used to design the coding and key of mobile network privacy big data. According to the two-dimensional coding characteristics of mobile network privacy big data in the key authorization protocol, the hybrid encryption and decryption key of mobile network privacy big data is designed, and the mixed encryption and decryption key of mobile network privacy big data is constructed, Realize the privacy of mobile network big data mixed encryption output and key design. The simulation results show that this method has good confidentiality and strong steganography performance, which improves the anti-attack ability of big data, which is used to encrypt the privacy of mobile network.

Nowadays, multimedia content retrieval has become the major service requirement of the Internet and the traffic of these contents has dominated the IP traffic. To reduce the duplicated traffic and improve the performance of distributing massive volumes of multimedia contents, in-network caching has been proposed recently. However, because in-network content caching can be directly utilized to respond users' requests, multimedia content retrieval is beyond content providers' control and makes it hard for them to implement access control and service accounting. In this paper, we propose an attribute-based accountable access control scheme for multimedia content distribution while making the best of in-network caching, in which content providers can be fully offline. In our scheme, the attribute-based encryption at multimedia content provider side and access policy based authentication at the edge router side jointly ensure the secure access control, which is also efficient in both space and time. Besides, secure service accounting is implemented by letting edge routers collect service credentials generated during users' request process. Through the informal security analysis, we prove the security of our scheme. Simulation results demonstrate that our scheme is efficient with acceptable overhead.

With the popularity of smart devices such as mobile phones and tablets, the security problem of the widely used ARMv8-A processor has received more and more attention. Flush+Reload and Flush+Flush cache attacks have become two of the most important security threats due to their low noise and high resolution. In order to resist Flush+Reload and Flush+Flush attacks, researchers proposed many defense methods. However, these existing methods have various shortcomings. The runtime defense methods using hardware performance counters cannot detect attacks fast enough, effectively detect Flush+Flush or avoid a high false positive rate. Static code analysis schemes are powerless for obfuscation techniques. The approaches of permanently reducing the resolution can only be utilized on browser products and cannot be applied in the system. In this paper, we design two more secure collaborative APIs-flush operation API and high resolution time API-which can resist Flush+Reload and Flush+Flush attacks. When the flush operation API is called, the high resolution time API temporarily reduces its resolution and automatically restores. Moreover, the flush operation API also has the ability to detect and handle suspected Flush+Reload and Flush+Flush attacks. The attack and performance comparison experiments prove that the two APIs we designed are safer and the performance losses are acceptable.

With the emergence of biometric technology in various applications, such as access control (e.g. mobile lock/unlock), financial transaction (e.g. Alibaba smile-to-pay) and time attendance, the development of biometric system attracts increasingly interest to the developers. Despite a sound biometric system gains the security assurance and great usability, it is a rather challenging task to develop an effective biometric system. For instance, many public available biometric APIs do not provide sufficient instructions / precise documentations on the usage of biometric APIs. Many developers are struggling in implementing these APIs in various tasks. Moreover, quick update on biometric-based algorithms (e.g. feature extraction and matching) may propagate to APIs, which leads to potential confusion to the system developers. Hence, we conduct an empirical study to the problems that the developers currently encountered while implementing the biometric APIs as well as the issues that need to be addressed when developing biometric systems using these APIs. We manually analyzed a total of 500 biometric API-related posts from various online media such as Stack Overflow and Neurotechnology. We reveal that 1) most of the problems encountered are related to the lack of precise documentation on the biometric APIs; 2) the incompatibility of biometric APIs cross multiple implementation environments.

K-anonymity is a popular model used in microdata publishing to protect individual privacy. This paper introduces the idea of ball tree and projection area density partition into k-anonymity algorithm.The traditional kd-tree implements the division by forming a super-rectangular, but the super-rectangular has the area angle, so it cannot guarantee that the records on the corner are most similar to the records in this area. In this paper, the super-sphere formed by the ball-tree is used to address this problem. We adopt projection area density partition to increase the density of the resulting recorded points. We implement our algorithm with the Gotrack dataset and the Adult dataset in UCI. The experimentation shows that the k-anonymity algorithm based on ball-tree and projection area density partition, obtains more anonymous groups, and the generalization rate is lower. The smaller the K is, the more obvious the result advantage is. The result indicates that our algorithm can make data usability even higher.

In this paper, we propose a conditional privacy-preserving authentication scheme based on pseudonyms and (t,n) threshold secret sharing, named CPPT, for vehicular communications. To achieve conditional privacy preservation, our scheme implements anonymous communications based on pseudonyms generated by hash chains. To prevent bad vehicles from conducting framed attacks on honest ones, CPPT introduces Shamir (t,n) threshold secret sharing technique. In addition, through two one-way hash chains, forward security and backward security are guaranteed, and it also optimize the revocation overhead. The size of certificate revocation list (CRL) is only proportional to the number of revoked vehicles and irrelated to how many pseudonymous certificates are held by the revoked vehicles. Extensive simulations demonstrate that CPPT outperforms ECPP, DCS, Hybrid and EMAP schemes in terms of revocation overhead, certificate updating overhead and authentication overhead.

This paper investigates the physical layer (PHY-layer) authentication that exploits channel state information (CSI) to enhance multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) system security by detecting spoofing attacks in wireless networks. A multi-user authentication system is proposed using convolutional neural networks (CNNs) which also can distinguish spoofers effectively. In addition, the mini batch scheme is used to train the neural networks and accelerate the training speed. Meanwhile, L1 regularization is adopted to prevent over-fitting and improve the authentication accuracy. The convolutional-neural-network-based (CNN-based) approach can authenticate legitimate users and detect attackers by CSIs with higher performances comparing to traditional hypothesis test based methods.

As the technological progress of mobile Internet, smartphone based on Android OS accounts for the vast majority of market share. The traditional encryption technology cannot resolve the dilemma in smartphone information leakage, and the Android-based authentication system in view of biometric recognition emerge to offer more reliable information assurance. In this paper, we summarize several biometrics providing their attributes. Furthermore, we also review the algorithmic framework and performance index acting on authentication techniques. Thus, typical identity authentication systems including their experimental results are concluded and analyzed in the survey. The article is written with an intention to provide an in-depth overview of Android-based biometric verification systems to the readers.

With the development of artificial intelligence (AI), multimedia privacy issues have become more challenging than ever. AI-assisted malicious entities can steal private information from multimedia data more easily than humans. Traditional multimedia privacy protection only considers the situation when humans are the adversaries, therefore they are ineffective against AI-assisted attackers. In this paper, we develop a new framework and new algorithms that can protect image privacy from both humans and AI. We combine the idea of adversarial image perturbation which is effective against AI and the obfuscation technique for human adversaries. Experiments show that our proposed methods work well for all types of attackers.

Although transport agency has employed desensitization techniques to deal with the privacy information when publicizing vehicle license plate recognition (VLPR) data, the adversaries can still eavesdrop on vehicle trajectories by certain means and further acquire the associated person and vehicle information through background knowledge. In this work, a privacy attacking method by using the desensitized VLPR data is proposed to link the vehicle trajectory. First the road average speed is evaluated by analyzing the changes of traffic flow, which is used to estimate the vehicle's travel time to the next VLPR system. Then the vehicle suspicion list is constructed through the time relevance of neighboring VLPR systems. Finally, since vehicles may have the same features like color, type, etc, the target trajectory will be located by filtering the suspected list by the rule of qualified identifier (QI) attributes and closest time method. Based on the Foshan City's VLPR data, the method is tested and results show that correct vehicle trajectory can be linked, which proves that the current VLPR data publication way has the risk of privacy disclosure. At last, the effects of related parameters on the proposed method are discussed and effective suggestions are made for publicizing VLPR date in the future.

In order to solve the interference of burrs, aliasing and other noises in the background area of millimeter wave human security inspection on the objects identification, an adaptive template matching filtering method is proposed. First, the preprocessed original image is segmented by level set algorithm, then the result is used as a template to filter the background of the original image. Finally, the image after background filtered is used as the input of bilateral filtering. The contrast experiments based on the actual millimeter wave image verifies the improvement of this algorithm compared with the traditional filtering method, and proves that this algorithm can filter the background noise of the human security image, retain the image details of the human body area, and is conducive to the object recognition and location in the millimeter wave security image.

At present, the AE method has the advantages of live measurement, online monitoring and easy fault location, so it is very suitable for insulation defect detection of power equipments such as GIS, etc. In this paper, development of a data processing software for PD acoustic detection based on a general fast analysis algorithm model is introduced. With considering the signal flow chart of current acoustic detection system widely used in operation and maintenance of power system equipments, the main function of the developed PD AE signals analysis software was designed, including the detailed analysis of individual data file, identification with phase compensation based on 2D PRPD histograms, batch processing analysis of data files, management of discharge fingerprint library and display of typical defect discharge data. And all of the corresponding developed software pages are displayed.

At present, the acoustic emission (AE) method has the advantages of live measurement and easy fault location, so it is very suitable for insulation defect detection of power equipments such as GIS, etc. While the conventional AE detection system or instruments always can't give a right discrimination result, because them always work based on the reference voltage or phase information from an auxiliary 220V voltage signal source rather than the operation high voltage (HV) with the real phase information corresponding to the detected AE pulsed signals. So there is a random phase difference between the reference phase and operation phase. The discharge fingerprint formed by the detected AE pulsed signals with reference phase using the same processing process is compared to the discharge fingerprint database formed in the HV laboratory with the real phase information, therefore, the system may not be able to discriminate the discharge mode of the field measured data from GIS in substation operation. In this paper, in order to design and develop a general fast analysis algorithm model for PD acoustic detection system to make an assistant diagnosis, the pattern identification with phase compensation was designed and applied. The results show that the method is effective and useful to deatl with AE signals meased in operation situation.

The characteristics of decentralization, tamper-resistance and transaction anonymity of blockchain can resolve effectively the problems in traditional intellectual property such as the difficulty of electronic obtaining for evidence, the high cost and low compensation when safeguarding the copyrights. Blockchain records the information through encryption algorithm, removes the third party, and stores the information in all nodes to prevent the information from being tampered with, so as to realize the protection of intellectual property. Starting from the bottom layer of blockchain, this paper expounds in detail the characteristics and the technical framework of blockchain. At the same time, according to the existing problems in transaction throughput, time delay and resource consumption of blockchain system, optimization mechanisms such as cross-chain and proof of stake are analyzed. Finally, combined with the characteristics of blockchain technology and existing application framework, this paper summarizes the existing problems in the industry and forecasts the development trend of intellectual property based on blockchain technology.

This paper presents a scheme of intellectual property protection of hardware circuit based on digital compression coding technology. The aim is to solve the problem of high embedding cost and low resource utilization of IP watermarking. In this scheme, the watermark information is preprocessed by dynamic compression coding around the idle circuit of FPGA, and the free resources of the surrounding circuit are optimized that the IP watermark can get the best compression coding model while the extraction and detection of IP core watermark by activating the decoding function. The experimental results show that this method not only expands the capacity of watermark information, but also reduces the cost of watermark and improves the security and robustness of watermark algorithm.

Virtual network function provides tenant with flexible and scalable end-to-end service chaining in the cloud computing and data center environments. However, comparing with traditional hardware network devices, the uncertainty caused by software and virtualization of Network Function Virtualization expands the attack surface, making the network node vulnerable to a certain types of attacks. The existing approaches for solving the problem of reliability are able to reduce the impact of failure of physical devices, but pay little attention to the attack scenario, which could be persistent and covert. In this paper, a heterogeneous backup strategy is brought up, enhancing the intrusion tolerance of NFV SFC by dynamically switching the VNF executor. The validity of the method is verified by simulation and game theory analysis.

Ciphertext storage can effectively solve the security problems in cloud storage, among which the ciphertext policy attribute-based encryption (CP-ABE) is more suitable for ciphertext access control in cloud storage environment for it can achieve one-to-many ciphertext sharing. The existing attribute encryption scheme CP-ABE has problems with revocation such as coarse granularity, untimeliness, and low efficiency, which cannot meet the demands of cloud storage. This paper proposes an RCP-ABE scheme that supports real-time revocable fine-grained attributes for the existing attribute revocable scheme, the scheme of this paper adopts the version control technology to realize the instant revocation of the attributes. In the key update mechanism, the subset coverage technology is used to update the key, which reduces the workload of the authority. The experimental analysis shows that RCP-ABE is more efficient than other schemes.

For future Internet, information-centric networking (ICN) is considered a potential solution to many of its current problems, such as content distribution, mobility, and security. Named Data Networking (NDN) is a more popular ICN project. However, concern regarding the protection of user data persists. Information caching in NDN decouples content and content publishers, which leads to content security threats due to lack of secure controls. Therefore, this paper presents a CP-ABE (ciphertext policy attribute based encryption) access control scheme based on hash table and data segmentation (CHTDS). Based on data segmentation, CHTDS uses a method of linearly splitting fixed data blocks, which effectively improves data management. CHTDS also introduces CP-ABE mechanism and hash table data structure to ensure secure access control and privilege revocation does not need to re-encrypt the published content. The analysis results show that CHTDS can effectively realize the security and fine-grained access control in the NDN environment, and reduce communication overhead for content access.